NSAIDs are promising chemopreventive agents in colorectal cancers, however their chemopreventive effectiveness is partial. Defining NSAIDs' chemopreventive mechanisms is important to identify molecular targets that can lead to develop better chemopreventive agents. NSAIDs induction of apoptosis is crucial to their antitumorigenic effects. We have found in previous studies that (a) hydroxyoctadecadienoic acid (13-SHODE), a linoleic acid product, and its producing enzyme, 15-lipoxygenase-1 (15-LOX-1), are downregulated in human colorectal cancers; and 13-S-HODE restores apoptosis in colorectal cancer cells, (b) NSAIDs downregulate GATA-6 expression to transcriptionally restore 15-LOX-1 expression that triggers apoptosis in human colorectal cancer cells, and (c) 15-LOX-1 downregulates the peroxisome proliferatoractivated receptor (PPAR)- delta (PPAR-delta) to trigger apoptosis. To test whether these chemopreventive mechanisms of NSAIDs are clinically relevant; our proposal will examine the following hypothesis: NSAIDs downregulate GATA-6 expression to restore 15-LOX-1 expression, which increases 13-S-HODE production to downregulate PPAR-delta expression, thereby inducing apoptosis in human colorectal tumors. This hypothesis will be tested in a clinical study of celecoxib treatment of patients with familial adenomatous polyposis syndrome (FAP) as follows: Specific Aim 1: To determine whether celecoxib downregulates GATA-6 expression to upregulate 15-LOX-1 expression, which in turn downregulates PPAR-d expression to induce apoptosis in human colorecta! polyps, we will measure GATA-6, 15-LOX-1 and PPAR-delta expression, 13-S-HODE levels, and apoptosis rates in colorectal polyps before and after 6 months of treatment with celecoxib in patients with familial adenomatous polyposis syndrome (FAP). Specific Aim 2: To define the relationship between chemopreventive response to celecoxib in patients with FAP' and the effects of celecoxib on 15-LOX-1 in human colorectal polyps, we will measure the response to celecoxib in terms of polyp number before and after 6 months of treatment with celecoxib in patients with FAP and will correlate the chemopreventive response (defined by the changes in mean polyp number) to 13-S-HODE levels before and after celecoxib treatment. If our hypothesis is confirmed, efforts can be directed to develop chemopreventive interventions that specifically target the GATA-6 and 15-LOX-1 signaling pathway.